Vol-Targeted Position Sizer · Intermediate
Free to readSizing to a Volatility Target
The core idea behind the Position Sizer in one page: why inverse-vol scaling equalizes risk contribution, and what happens to portfolio volatility when you apply it consistently.
You can size a trade by the number of shares you "feel" like owning, by a fixed dollar amount, or by the volatility that position will actually contribute to your portfolio. Only the third approach delivers portfolios where every position counts equally. This page explains the mechanics of volatility targeting and why it's a structural improvement over the alternatives.
The problem with fixed-dollar sizing
Suppose you put $5,000 into each of three positions: a utility ETF with 12% realized vol, an equity ETF with 20% realized vol, and a single biotech with 60% realized vol. On paper you have three equal positions. In practice the biotech contributes 25 times the daily variance of the utility — it is almost the entire risk of the portfolio. If it blows up, the other two names were decorations. This is the silent flaw in fixed-dollar sizing: it confuses equal capital with equal risk, and they are rarely the same thing.
Fixed-dollar sizing (left): the high-vol name towers over the others in variance contribution. Vol-targeted sizing (right): each name contributes the same expected variance, so no single position can dominate the portfolio's risk.
The inverse-vol formula
The math is straightforward. If your per-position volatility target is T and the underlying's annualized realized vol is σ, the fraction of capital to allocate is T / σ. A 1% per-position target on a 20%-vol name gives 5% of capital; on a 40%-vol name, 2.5%. The high-vol name gets half the notional — and delivers the same expected risk contribution. Scale that across a book of uncorrelated positions and portfolio vol stays predictable rather than dominated by whoever happened to be the wildest name last week.
How the regime multiplier changes the picture
The inverse-vol formula gives you the "full size" for a given regime. But the regime multiplier from Layer 1 scales it back when conditions are stressed. If the underlying's realized vol is in its 90th percentile — historically extreme — the tool applies a multiplier that can cut the calculated size nearly in half. This is not a contradiction of vol-targeting; it's an acknowledgment that realized vol in extreme percentiles predicts continued elevated vol, so "today's vol" understates the risk of the next few days. The multiplier is a lookback correction on top of the point-in-time scaling.
Why this smooths portfolio volatility
When realized vol spikes, a vol-targeted portfolio automatically sheds gross exposure — positions get smaller — before that spike has time to inflict maximum damage. When markets calm, size gradually restores. The portfolio's vol is not constant (market vol changes), but it is managed rather than accidental. Across a year of varied regimes, the result is a vol path that stays closer to the target and drawdowns that are less catastrophic because the largest bets were placed when conditions were calmest.
It is not a stop-loss. If a position moves against you before you can act, the loss can still exceed the vol-implied target. It's a sizing rule, not a guarantee. Correlation between positions also matters — ten "equal-vol" positions in the same sector are not ten independent bets. The tool sizes each trade; you manage the book.
The concept is free. To calculate a live vol-targeted size on your names: ETFs via ETF Analytics, single stocks via ETF + Equities. The full realized-vol percentile history and regime multiplier table are available with Everything.
See plans →Educational content from Nations Indexes. VolDex® and RiskDex® are registered marks of Nations Indexes. Diagrams are schematic. Nothing here is investment advice.